In the paper Monotonicity for Elliptic Equations in Unbounded Lipschitz Domains - Berestycki - Caffarelli - Niremberg, they use the estimate "obtained by interior elliptic regularity": if $B=B_\rho(y)\subset\Omega$ and $|x-y|\leq\rho/2$, then $$|u(x)-u(y)|\leq \widehat C\max_Bu\cdot\left(\frac{|x-y|}{\rho}\right)^\gamma+\widehat C\rho^{2-\gamma}|x-y|^\gamma,$$ for $\gamma\in(0,1]$. Someone can help me to obtain this estimate? Thank you.


I don't know why they write $\gamma$ there. Observe that $\gamma\to a^\gamma$ is a decreasing function of $\gamma$ for any fixed base $0<a<1$. Here we have $$ \frac{|x-y|}{\rho}\le \frac12 $$ and since the right hand side is $$ \widehat C\max_Bu\cdot\left(\frac{|x-y|}{\rho}\right)^\gamma+\widehat C\rho^{2}\left(\frac{|x-y|}{\rho}\right)^\gamma $$ it is a decreasing function of $\gamma$. Thus, it suffices to prove the estimate for $\gamma=1$. Theorem 3.9 in Gilbarg-Trudinger (used with $\Omega=B$) gives the gradient bound $$ \sup_{B_{\rho/2}(y)} |\nabla u| \le \frac{C}{\rho}\left( \sup_B |u| + \rho^2 \sup_B |\Delta u|\right) $$ Recalling that $\Delta u$ is uniformly bounded, and using the mean value theorem, we obtain $$ |u(x)-y(y)| \le \frac{C}{\rho}|x-y|\left( \sup_B |u| + \rho^2 \sup_B |\Delta u|\right) $$ for $x\in B_{\rho/2}(y)$.

  • $\begingroup$ The $\gamma$ is used to prove that $u$ is Hölder Continuous. In the paper, we use this estimate to prove this. Can I use this estimate argument for each $\gamma\in(0,1]$? 5PM, you answered an another question to me here: math.stackexchange.com/questions/284506/… but I have one more doubt, about the uniform convergence of the Laplacian (the $C^2$-convergence). Do you know what implies this convergence? Thank you very much!!! $\endgroup$ – José Carlos Feb 4 '13 at 18:35
  • $\begingroup$ I didn't understand why the right hand side is a decreasing function of $\gamma$. Could you explain it? $\endgroup$ – José Carlos Feb 4 '13 at 19:29
  • $\begingroup$ @JoséCarlos I updated my answer with more detail. I'll get back to the convergence question.. $\endgroup$ – user53153 Feb 4 '13 at 19:52
  • $\begingroup$ Thank you very much @5PM. $\endgroup$ – José Carlos Feb 4 '13 at 19:53
  • $\begingroup$ I am waiting your help in the other question @5PM! Thank you! $\endgroup$ – José Carlos Feb 4 '13 at 20:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.